def on_deleted(self, harvest_object, header):
""" See :meth:`OAIPMHHarvester.on_deleted`
Mark package for deletion.
"""
package_id = get_package_id_by_pid(header.identifier(), 'primary')
if package_id:
harvest_object.package_id = package_id
harvest_object.content = None
harvest_object.report_status = "deleted"
harvest_object.save()
return True
def test_get_package_id_by_pid(self):
package_1_id, package_2_id = self._create_datasets()
self.assertEquals(utils.get_package_id_by_pid('some_primary_pid_1', 'primary'), package_1_id)
self.assertEquals(utils.get_package_id_by_pid('some_metadata_pid_1', 'relation'), package_1_id)
self.assertEquals(utils.get_package_id_by_pid('some_data_pid_1', 'unknown_type'), None)
self.assertEquals(utils.get_package_id_by_pid('some_data_pid_2', 'relation'), package_2_id)
self.assertEquals(utils.get_package_id_by_pid('some_part_pid_2', 'relation'), package_2_id)
self.assertEquals(utils.get_package_id_by_pid('some_unknown_pid_2', 'relation'), None)
self.assertEquals(utils.get_package_id_by_pid('invalid', 'invalid'), None)
def test_get_package_id_by_pid(self):
package_1_id, package_2_id = self._create_datasets()
self.assertEquals(utils.get_package_id_by_pid('some_data_pid_1', 'data'), package_1_id)
self.assertEquals(utils.get_package_id_by_pid('some_metadata_pid_1', 'metadata'), package_1_id)
self.assertEquals(utils.get_package_id_by_pid('some_data_pid_1', 'metadata'), None)
self.assertEquals(utils.get_package_id_by_pid('some_data_pid_2', 'data'), package_2_id)
self.assertEquals(utils.get_package_id_by_pid('some_version_pid_2', 'version'), package_2_id)
self.assertEquals(utils.get_package_id_by_pid('some_version_pid_2', 'data'), None)
self.assertEquals(utils.get_package_id_by_pid('invalid', 'version'), None)
def read_data(self, xml):
""" Extract package data from given XML.
:param xml: xml element (lxml)
:return: dictionary
"""
cmd = first(
xml.xpath('//oai:record/oai:metadata/cmd:CMD',
namespaces=self.namespaces))
if cmd is None:
raise CmdiReaderException(
"Unexpected XML format: No CMD -element found")
resource_info = cmd.xpath("//cmd:Components/cmd:resourceInfo",
namespaces=self.namespaces)[0]
if resource_info is None:
raise CmdiReaderException(
"Unexpected XML format: No resourceInfo -element found")
metadata_identifiers = self._text_xpath(
cmd, "//cmd:identificationInfo/cmd:identifier/text()")
languages = self._text_xpath(
cmd,
"//cmd:corpusInfo/cmd:corpusMediaType/cmd:corpusTextInfo/cmd:languageInfo/cmd:languageId/text()"
)
# convert the descriptions to a JSON string of type {"fin":"kuvaus", "eng","desc"}
desc_json = {}
for desc in xml.xpath("//cmd:identificationInfo/cmd:description",
namespaces=self.namespaces):
lang = convert_language(
desc.get('{http://www.w3.org/XML/1998/namespace}lang',
'undefined').strip())
desc_json[lang] = unicode(desc.text).strip()
description = json.dumps(desc_json)
# convert the titles to a JSON string of type {"fin":"otsikko", "eng","title"}
transl_json = {}
for title in xml.xpath('//cmd:identificationInfo/cmd:resourceName',
namespaces=self.namespaces):
lang = convert_language(
title.get('{http://www.w3.org/XML/1998/namespace}lang',
'undefined').strip())
transl_json[lang] = title.text.strip()
title = json.dumps(transl_json)
provider = self.provider
version = first(
self._text_xpath(
resource_info,
"//cmd:metadataInfo/cmd:metadataLastDateUpdated/text()")) or ""
coverage = first(
self._text_xpath(
resource_info,
"//cmd:corpusInfo/cmd:corpusMediaType/cmd:corpusTextInfo/cmd:timeCoverageInfo/cmd:timeCoverage/text()"
)) or ""
pids = []
primary_pid = ''
direct_download_URL = ''
access_request_URL = ''
access_application_URL = ''
# data_identifiers = self._text_xpath(cmd, "//cmd:identificationInfo/cmd:url/text()")
for pid in [
CmdiReader._language_bank_urn_pid_enhancement(metadata_pid)
for metadata_pid in metadata_identifiers
]:
if 'urn' in pid and not primary_pid:
pids.append(dict(id=pid, provider=provider, type='primary'))
primary_pid = pid
# else:
# pids.append(dict(id=pid, provider=provider, type='relation', relation='generalRelation'))
#
# pids += [dict(id=CmdiReader._language_bank_urn_pid_enhancement(pid), provider=provider, type='relation',
# relation='generalRelation') for pid in data_identifiers]
license_identifier = CmdiReader._language_bank_license_enhancement(
first(
self._text_xpath(
resource_info,
"//cmd:distributionInfo/cmd:licenceInfo/cmd:licence/text()"
)) or 'notspecified')
availability = CmdiReader._language_bank_availability_from_license(
license_identifier)
if license_identifier.lower().strip() != 'undernegotiation':
if availability == 'direct_download':
direct_download_URL = primary_pid
if availability == 'access_request':
access_request_URL = primary_pid
if availability == 'access_application_other':
sliced_pid = primary_pid.rsplit('/', 1)
if len(sliced_pid) >= 2:
access_application_URL = 'https://lbr.csc.fi/web/guest/catalogue?domain=LBR&target=basket&resource=' + sliced_pid[
1]
temporal_coverage_begin = ""
temporal_coverage_end = ""
if coverage:
split = [item.strip() for item in coverage.split("-")]
if len(split) == 2:
temporal_coverage_begin = split[0]
temporal_coverage_end = split[1]
# TODO: Check agent mapping.
#print "###", _get_persons(resource_info, "//cmd:distributionInfo/cmd:licenceInfo/cmd:licensorPerson")
#print "###", _get_persons(resource_info, "//cmd:distributionInfo/cmd:licenceInfo/cmd:distributionRightsHolderPerson")
#print "###", _get_persons(resource_info, "//cmd:distributionInfo/cmd:iprHolderPerson")
#print "###", _get_persons(resource_info, "//cmd:contactPerson")
#print "###", _get_persons(resource_info, "//cmd:metadataInfo/cmd:metadataCreator")
#print "###", _get_organizations(resource_info, "//cmd:distributionInfo/cmd:licenceInfo/cmd:licensorOrganization")
#print "###", _get_organizations(resource_info, "//cmd:distributionInfo/cmd:licenceInfo/cmd:distributionRightsHolderOrganization")
#print "###", _get_organizations(resource_info, "//cmd:distributionInfo/cmd:iprHolderOrganization")
contacts = self._persons_as_contact(
self._get_persons(resource_info, "//cmd:contactPerson"))
agents = []
agents.extend(
self._persons_as_agent(
self._get_persons(
resource_info,
"//cmd:distributionInfo/cmd:iprHolderPerson"), 'author'))
agents.extend(
self._persons_as_agent(
self._get_persons(
resource_info,
"//cmd:distributionInfo/cmd:licenceInfo/cmd:distributionRightsHolderPerson"
), 'owner'))
agents.extend(
self._organization_as_agent(
self._get_organizations(
resource_info,
"//cmd:distributionInfo/cmd:iprHolderOrganization"),
'author'))
agents.extend(
self._organization_as_agent(
self._get_organizations(
resource_info,
"//cmd:distributionInfo/cmd:licenceInfo/cmd:distributionRightsHolderOrganization"
), 'owner'))
existing_package_id = get_package_id_by_pid(primary_pid, u'primary')
package_id = existing_package_id if existing_package_id else get_unique_package_id(
)
result = {
'name': pid_to_name(package_id),
'language': ",".join(languages),
'pids': pids,
'version': version,
'notes': description,
'title': title,
'type': 'dataset',
'contact': contacts,
'agent': agents,
'availability': availability,
'direct_download_URL': direct_download_URL,
'access_request_URL': access_request_URL,
'access_application_URL': access_application_URL,
'temporal_coverage_begin': temporal_coverage_begin,
'temporal_coverage_end': temporal_coverage_end,
'license_id': license_identifier,
'license_URL': ''
}
if not languages:
result['langdis'] = u'True'
if package_id:
result['id'] = package_id
# TODO: Ask about distributionAccessMedium
# _strip_first(_text_xpath(resource_info, "//cmd:distributionInfo/availability/text()"))
# url = _strip_first(_text_xpath(resource_info, "//cmd:identificationInfo/cmd:url/text()"))
return result
def _ddi2ckan(self, original_url, original_xml, harvest_object):
'''Extract package values from bs4 object 'ddi_xml' parsed from xml
'''
# TODO: Use .extract() and .string.extract() function so handled elements are removed from ddi_xml.
doc_citation = "ddi_xml.codeBook.docDscr.citation"
stdy_dscr = "ddi_xml.codeBook.stdyDscr"
####################################################################
# Read mandatory metadata fields: #
####################################################################
# Authors & organizations
authors = self.get_authors(self.ddi_xml.stdyDscr.citation, 'AuthEnty')
agent = authors[:]
agent.extend(self.get_contributors(self.ddi_xml.stdyDscr.citation))
# Availability
availability = AVAILABILITY_DEFAULT
if _access_request_URL_is_found():
availability = 'direct_download'
if _is_fsd(original_url):
availability = AVAILABILITY_FSD
# Keywords
keywords = self.get_keywords(self.ddi_xml.stdyDscr.stdyInfo.subject)
# Language
# TODO: Where/how to extract multiple languages: 'language': u'eng, fin, swe' ?
language = self.convert_language(
self._read_value("ddi_xml.codeBook.get('xml:lang')"))
# Titles
titles = self._read_value(stdy_dscr + ".citation.titlStmt(['titl', 'parTitl'])") or \
self._read_value(doc_citation + ".titlStmt(['titl', 'parTitl'])", mandatory_field=True)
# langtitle=[dict(lang=self.convert_language(a.get('xml:lang', '')), value=a.text) for a in titles]
# [{"lang":"fin", "value":"otsikko"}, {"lang:"en", "value":"title"}]
# convert the titles to a JSON string of type {"fin":"otsikko", "eng","title"}
transl_json = {}
first_title = ""
# default to finnish, since first title has no lang value, which causes the validator to whine
# we might want to update the DDI harvester to accept a language configuration parameter, if
# we decide to harvest DDI resources from other sources.
default_lang = "fi"
for title in titles:
transl_json[self.convert_language(title.get('xml:lang', default_lang))] = title.text
# we want to get save the first title for use lateron
if not first_title:
first_title = title.text
title = json.dumps(transl_json)
# License
# TODO: Extract prettier output. Should we check that element contains something?
# Should this be in optional section if not mandatory_field?
license_url = self._read_value(stdy_dscr + ".dataAccs.useStmt.get_text(separator=u' ')", mandatory_field=False)
if _is_fsd(original_url):
license_id = LICENSE_ID_FSD
else:
license_id = LICENSE_ID_DEFAULT
# Contact (package_extra.key: contact_[k]_name in database, contact in WUI)
contact_name = self._read_value(stdy_dscr + ".citation.distStmt('contact')") or \
self._read_value(stdy_dscr + ".citation.distStmt('distrbtr')") or \
self._read_value(doc_citation + ".prodStmt('producer')", mandatory_field=True)
# TODO: clean out (or ask FSD to clean) mid text newlines (eg. in FSD2482)
if contact_name and contact_name[0].text:
contact_name = contact_name[0].text
else:
contact_name = self._read_value(stdy_dscr + ".citation.prodStmt.producer.get('affiliation')", mandatory_field=True)
if _is_fsd(original_url):
contact_email = CONTACT_EMAIL_FSD
# TODO: Allow trying other email also in FSD metadata
else:
contact_email = self._read_value(stdy_dscr + ".citation.distStmt.contact.get('email')", mandatory_field=True)
# Modified date
version = self.get_attr_optional(self.ddi_xml.stdyDscr.citation,
'prodDate', 'date') or \
self.get_attr_mandatory(self.ddi_xml.stdyDscr.citation,
'version', 'date')
# This idNos is an FSD specific solution
idNos = self._read_value(stdy_dscr + ".citation.titlStmt.find_all('IDNo')", mandatory_field=False)
if not idNos:
idNos = self._read_value(doc_citation + ".titlStmt.find_all('IDNo')", mandatory_field=True)
pids = list()
idNoValues = [bsIdNo.text for bsIdNo in idNos]
agencies = [bsIdNo.get('agency') for bsIdNo in idNos]
primary_pid = None
if len(idNoValues) == len(agencies):
for idNoVal, agency in zip(idNoValues, agencies):
if agency == 'Kansalli' \
'skirjasto':
pids.append({'id': idNoVal, 'type': 'primary', 'provider': agency})
primary_pid = idNoVal
else:
pids.append({'id': agency + idNoVal, 'type': 'relation', 'provider': agency, 'relation': 'generalRelation'})
# Should we generate a version PID?
# vpid = utils.generate_pid()
# pids.append({'id': vpid, 'type': 'version', 'provider': 'kata'})
# Original web page as resource
# For FSD 'URI' leads to summary web page of data, hence format='html'
orig_web_page = self._read_value(doc_citation + ".holdings.get('URI', '')")
if orig_web_page:
orig_web_page_resource = {'description': first_title,
'format': u'html',
'resource_type': 'documentation',
'url': orig_web_page}
else:
orig_web_page_resource = {}
# Owner
owner = self._read_value(stdy_dscr + ".citation.prodStmt.producer.text") or \
self._read_value(stdy_dscr + ".citation.rspStmt.AuthEnty.text") or \
self._read_value(doc_citation + ".prodStmt.producer.string", mandatory_field=True)
agent.append({'role': 'owner',
'name': owner})
# Owner organisation
if harvest_object:
hsid = harvest_object.harvest_source_id
hsooid = model.Session.query(model.Package).filter(model.Package.id==hsid).one().owner_org
owner_org = model.Session.query(model.Group).filter(model.Group.id==hsooid).one().name
else:
owner_org = u''
# Distributor (Agent: distributor, the same is used as contact)
agent.append({
'role': 'distributor',
'name': contact_name})
####################################################################
# Read optional metadata fields: #
####################################################################
# Availability
if _is_fsd(original_url):
access_request_url = ACCESS_REQUEST_URL_FSD
else:
access_request_url = u''
# Contact
contact_phone = self._read_value(doc_citation + ".holdings.get('callno')") or \
self._read_value(stdy_dscr + ".citation.holdings.get('callno')")
contact_URL = self._read_value( stdy_dscr + ".dataAccs.setAvail.accsPlac.get('URI')") or \
self._read_value( stdy_dscr + ".citation.distStmt.contact.get('URI')") or \
self._read_value( stdy_dscr + ".citation.distStmt.distrbtr.get('URI')") or \
CONTACT_URL_FSD if _is_fsd(original_url) else None
# convert the descriptions to a JSON string of type {"fin":"aineiston kuvaus", "eng","dataset description"}
descriptions = self._read_value(stdy_dscr + ".stdyInfo.abstract('p')")
if not descriptions:
descriptions = self._read_value(stdy_dscr + ".citation.serStmt.serInfo('p')")
translated_notes = {}
for des in descriptions:
lang = self.convert_language(des.get('xml:lang', 'fi'))
if lang in translated_notes:
translated_notes[lang] += '\r\n\r\n' + des.text
else:
translated_notes[lang] = des.text
notes = json.dumps(translated_notes)
# Discipline
discipline = self.get_discipline(self.ddi_xml.stdyDscr.stdyInfo.subject)
# Dataset lifetime events
events = self._get_events(stdy_dscr, authors)
# Geographic coverage
geo_cover = self.get_geo_coverage(self.ddi_xml)
# Temporal coverage
temp_start, temp_end = self.get_temporal_coverage(self.ddi_xml)
# Citation
citation = self._read_value(stdy_dscr + ".citation.biblCit.text", mandatory_field=False)
####################################################################
# Flatten rest to 'XPath/path/to/element': 'value' pairs #
####################################################################
etree_xml = etree.fromstring(str(self.ddi_xml))
flattened_ddi = importcore.generic_xml_metadata_reader(etree_xml.find('.//{*}docDscr'))
xpath_dict = flattened_ddi.getMap()
flattened_ddi = importcore.generic_xml_metadata_reader(etree_xml.find('.//{*}stdyDscr'))
xpath_dict.update(flattened_ddi.getMap())
existing_package_id = get_package_id_by_pid(primary_pid, u'primary')
package_id = existing_package_id if existing_package_id else get_unique_package_id()
package_name = pid_to_name(package_id)
package_dict = dict(
access_application_URL=u'',
access_request_URL=unicode(access_request_url),
agent=agent,
algorithm=u'', # To be implemented straight in 'resources'
availability=unicode(availability),
contact=[{'name': contact_name,
'email': contact_email,
'URL': contact_URL,
'phone': contact_phone}],
direct_download_URL=u'', # To be implemented straight in 'resources
discipline=discipline,
event=events,
geographic_coverage=geo_cover,
groups=[],
id=package_id,
langdis=u'True', # HUOMAA!
language=language,
license_URL=license_url,
license_id=license_id,
mimetype=u'', # To be implemented straight in 'resources
name=package_name,
notes=notes or u'',
pids=pids,
owner_org=owner_org,
resources=[orig_web_page_resource],
tag_string=keywords,
temporal_coverage_begin=temp_start,
temporal_coverage_end=temp_end,
title=title,
type='dataset',
version=version,
version_PID='',
citation=citation
)
package_dict['xpaths'] = xpath_dict
# Above line creates:
# package_dict = {
# 'access_request_url': 'some_url',
# # ...
# 'xpaths': {'stdyDscr/othrStdyMat.0/relPubl.34':
# 'Uskon asia: nuorisobarometri 2006 (2006).'},
# {'stdyD...': 'Some value'}]
# }
#package_dict['extras'].update(_save_ddi_variables_to_csv(ddi_xml, somepkg))
# Vanhojen koodien järjestys:
#_save_original_xml_and_link_as_resources()
#_save_ddi_variables_to_csv()
#_create_group_based_on_organizations()
#_last_statements_to_rewrite()
# JuhoL: Set harvest object to some end state and commit
if harvest_object is not None:
harvest_object.content = None
# Should this be flushed? model.Session.flush()
#model.repo.commit()
return package_dict
def read_data(self, xml):
""" Extract package data from given XML.
:param xml: xml element (lxml)
:return: dictionary
"""
# MAP DATACITE MANDATORY FIELD
# Identifier to primary pid
identifier = xml.find('.//{http://datacite.org/schema/kernel-3}identifier')
pids = [{
'id': identifier.text,
'type': 'primary',
'provider': identifier.get('identifierType')}]
# Creator name to agent
# TODO: map nameIdentifier to agent.id and nameIdentifierScheme and schemeURI
# to extras
agents = []
for creator in xml.findall('.//{http://datacite.org/schema/kernel-3}creator'):
creatorName = creator.find('.//{http://datacite.org/schema/kernel-3}creatorName').text
creatorAffiliation = creator.find('.//{http://datacite.org/schema/kernel-3}affiliation').text
agents.append({
'role': u'author',
'name': creatorName,
'organisation': creatorAffiliation
})
# Primary title to title
# TODO: if titleType is present, check to find out if title is actually primary
# TODO: map non-primary titles to extras
title = xml.find('.//{http://datacite.org/schema/kernel-3}title').text
langtitle = [{'lang': 'en', 'value': title}] # Assuming we always harvest English
# Publisher to contact
publisher = xml.find('.//{http://datacite.org/schema/kernel-3}publisher').text
contacts = [{'name': publisher}]
# Publication year to event
publication_year = xml.find('.//{http://datacite.org/schema/kernel-3}publicationYear').text
events = [{'type': u'published', 'when': publication_year, 'who': publisher, 'descr': u'Dataset was published'}]
# MAP DATACITE RECOMMENDED FIELDS
# Subject to tags
# TODO: map subjectsScheme and schemeURI to extras
# Contributor to agent
# TODO: map nameIdentifier to agent.id, nameIdentifierScheme, schemeURI and
# contributorType to extras
for contributor in xml.findall('.//{http://datacite.org/schema/kernel-3}contributor'):
contributorName = contributor.find('.//{http://datacite.org/schema/kernel-3}contributorName').text
contributorAffiliation = contributor.find('.//{http://datacite.org/schema/kernel-3}affiliation').text
agents.append({
'role': u'contributor',
'name': contributorName,
'organisation': contributorAffiliation
})
# Date to event
for date in xml.findall('.//{http://datacite.org/schema/kernel-3}date'):
events.append({
'type': date.get('dateType'),
'when': date.text,
'who': u'unknown',
'descr': date.get('dateType'),
})
# ResourceType to extra
# TODO: map resourceType and resourceTypeGeneral to extras
# RelatedIdentifier to showcase
# TODO: map RelatedIdentifier to showcase title, relatedIdentifierType, relationType,
# relatedMetadataScheme, schemeURI and schemeType to showcase description
# Description to langnotes
description = ''
for element in xml.findall('.//{http://datacite.org/schema/kernel-3}description'):
description += element.get('descriptionType') + ': ' + element.text + ' '
langnotes = [{
'lang': 'en', # Assuming we always harvest English
'value': description,
}]
# GeoLocation to geograhic_coverage
# TODO: map geoLocationPoint and geoLocationBox to extras, geoLocationPlace to
# geographic_coverage
# MAP DATACITE OPTIONAL FIELDS
# Language to language
# TODO: map language to language
# AlternateIdentifier to pids
# TODO: map AlternateIdentifier to pids.id, alternateIdentifierType to pids.provider
# Size to extra
# TODO: map size to extra
# Format to resources
# TODO: map format to resources.format
# Version to extra
# DataCite version is a string such as 'v3.2.1' and can't be used as Etsin version
# TODO: map version to extra
# Rights to license
license_URL = ''
for right in xml.findall('.//{http://datacite.org/schema/kernel-3}rights'):
license_URL += right.text + ' ' + right.get('rightsURI') + ' '
# OTHER - REQUIRED BY CKANEXT-HARVEST
# Get or create package id
existing_package_id = get_package_id_by_pid(identifier.text, u'primary')
package_id = existing_package_id if existing_package_id else get_unique_package_id()
result = {
'agent': agents,
'contact': contacts,
'event': events,
'id': package_id,
'langnotes': langnotes,
'langtitle': langtitle,
'license_URL': license_URL,
'pids': pids,
'title': title,
'type': 'dataset',
'version': datetime.datetime.now().strftime("%Y-%m-%d")
}
return result
def read_data(self, xml):
""" Extract package data from given XML.
:param xml: xml element (lxml)
:return: dictionary
"""
cmd = first(xml.xpath('//oai:record/oai:metadata/cmd:CMD', namespaces=self.namespaces))
if cmd is None:
raise CmdiReaderException("Unexpected XML format: No CMD -element found")
resource_info = cmd.xpath("//cmd:Components/cmd:resourceInfo", namespaces=self.namespaces)[0]
if resource_info is None:
raise CmdiReaderException("Unexpected XML format: No resourceInfo -element found")
metadata_identifiers = self._text_xpath(cmd, "//cmd:identificationInfo/cmd:identifier/text()")
languages = self._text_xpath(cmd, "//cmd:corpusInfo/cmd:corpusMediaType/cmd:corpusTextInfo/cmd:languageInfo/cmd:languageId/text()")
# convert the descriptions to a JSON string of type {"fin":"kuvaus", "eng","desc"}
desc_json = {}
for desc in xml.xpath("//cmd:identificationInfo/cmd:description", namespaces=self.namespaces):
lang = convert_language(desc.get('{http://www.w3.org/XML/1998/namespace}lang', 'undefined').strip())
desc_json[lang] = unicode(desc.text).strip()
description = json.dumps(desc_json)
# convert the titles to a JSON string of type {"fin":"otsikko", "eng","title"}
transl_json = {}
for title in xml.xpath('//cmd:identificationInfo/cmd:resourceName', namespaces=self.namespaces):
lang = convert_language(title.get('{http://www.w3.org/XML/1998/namespace}lang', 'undefined').strip())
transl_json[lang] = title.text.strip()
title = json.dumps(transl_json)
provider = self.provider
version = first(self._text_xpath(resource_info, "//cmd:metadataInfo/cmd:metadataLastDateUpdated/text()")) or ""
coverage = first(self._text_xpath(resource_info, "//cmd:corpusInfo/cmd:corpusMediaType/cmd:corpusTextInfo/cmd:timeCoverageInfo/cmd:timeCoverage/text()")) or ""
pids = []
primary_pid = ''
direct_download_URL = ''
access_request_URL = ''
access_application_URL = ''
# data_identifiers = self._text_xpath(cmd, "//cmd:identificationInfo/cmd:url/text()")
for pid in [CmdiReader._language_bank_urn_pid_enhancement(metadata_pid) for metadata_pid in metadata_identifiers]:
if 'urn' in pid and not primary_pid:
pids.append(dict(id=pid, provider=provider, type='primary'))
primary_pid=pid
# else:
# pids.append(dict(id=pid, provider=provider, type='relation', relation='generalRelation'))
#
# pids += [dict(id=CmdiReader._language_bank_urn_pid_enhancement(pid), provider=provider, type='relation',
# relation='generalRelation') for pid in data_identifiers]
license_identifier = CmdiReader._language_bank_license_enhancement(first(self._text_xpath(resource_info, "//cmd:distributionInfo/cmd:licenceInfo/cmd:licence/text()")) or 'notspecified')
availability = CmdiReader._language_bank_availability_from_license(license_identifier)
if license_identifier.lower().strip() != 'undernegotiation':
if availability == 'direct_download':
direct_download_URL = primary_pid
if availability == 'access_request':
access_request_URL = primary_pid
if availability == 'access_application_other':
sliced_pid = primary_pid.rsplit('/', 1)
if len(sliced_pid) >= 2:
access_application_URL = 'https://lbr.csc.fi/web/guest/catalogue?domain=LBR&target=basket&resource=' + sliced_pid[1]
temporal_coverage_begin = ""
temporal_coverage_end = ""
if coverage:
split = [item.strip() for item in coverage.split("-")]
if len(split) == 2:
temporal_coverage_begin = split[0]
temporal_coverage_end = split[1]
# TODO: Check agent mapping.
#print "###", _get_persons(resource_info, "//cmd:distributionInfo/cmd:licenceInfo/cmd:licensorPerson")
#print "###", _get_persons(resource_info, "//cmd:distributionInfo/cmd:licenceInfo/cmd:distributionRightsHolderPerson")
#print "###", _get_persons(resource_info, "//cmd:distributionInfo/cmd:iprHolderPerson")
#print "###", _get_persons(resource_info, "//cmd:contactPerson")
#print "###", _get_persons(resource_info, "//cmd:metadataInfo/cmd:metadataCreator")
#print "###", _get_organizations(resource_info, "//cmd:distributionInfo/cmd:licenceInfo/cmd:licensorOrganization")
#print "###", _get_organizations(resource_info, "//cmd:distributionInfo/cmd:licenceInfo/cmd:distributionRightsHolderOrganization")
#print "###", _get_organizations(resource_info, "//cmd:distributionInfo/cmd:iprHolderOrganization")
contacts = self._persons_as_contact(self._get_persons(resource_info, "//cmd:contactPerson"))
agents = []
agents.extend(self._persons_as_agent(self._get_persons(resource_info, "//cmd:distributionInfo/cmd:iprHolderPerson"), 'author'))
agents.extend(self._persons_as_agent(self._get_persons(resource_info, "//cmd:distributionInfo/cmd:licenceInfo/cmd:distributionRightsHolderPerson"), 'owner'))
agents.extend(self._organization_as_agent(self._get_organizations(resource_info, "//cmd:distributionInfo/cmd:iprHolderOrganization"), 'author'))
agents.extend(self._organization_as_agent(self._get_organizations(resource_info, "//cmd:distributionInfo/cmd:licenceInfo/cmd:distributionRightsHolderOrganization"), 'owner'))
existing_package_id = get_package_id_by_pid(primary_pid, u'primary')
package_id = existing_package_id if existing_package_id else get_unique_package_id()
result = {'name': pid_to_name(package_id),
'language': ",".join(languages),
'pids': pids,
'version': version,
'notes': description,
'title': title,
'type': 'dataset',
'contact': contacts,
'agent': agents,
'availability': availability,
'direct_download_URL': direct_download_URL,
'access_request_URL': access_request_URL,
'access_application_URL': access_application_URL,
'temporal_coverage_begin': temporal_coverage_begin,
'temporal_coverage_end': temporal_coverage_end,
'license_id': license_identifier,
'license_URL': ''}
if not languages:
result['langdis'] = u'True'
if package_id:
result['id'] = package_id
# TODO: Ask about distributionAccessMedium
# _strip_first(_text_xpath(resource_info, "//cmd:distributionInfo/availability/text()"))
# url = _strip_first(_text_xpath(resource_info, "//cmd:identificationInfo/cmd:url/text()"))
return result